DVOA Analysis
Football Outsiders' revolutionary metrics that break down every single play of the NFL season

Week 18 DVOA Ratings

Ryan Tannehill
Photo: USA Today Sports Images

Once again, it is time for postseason DVOA ratings. As always, the following rules apply:

  • All 32 teams are ranked, whether they made the playoffs or not.
  • Teams are ranked in order of weighted DVOA, not total season DVOA. Since weighted DVOA is meant to lower the strength of older games, these ratings do not include Weeks 1-4, and Weeks 5-10 are somewhat discounted.
  • Only weighted DVOA is listed for offense, defense, and special teams. Total DVOA is also listed, but one game doesn't change much in a 17-game sample so these ratings will be similar to those from the end of the season.
  • Teams which did not play in the wild-card round are treated as if they had a bye week. (That includes both the 20 non-playoff teams and the four teams with byes.) However, this year for the playoff odds report, we used ratings as of the end of Week 17 for these four teams rather than using the ratings listed below.

The Baltimore Ravens' rating at No. 1 is now really, really, absurdly high. We're up to 55.7%. Both of Baltimore's early-season losses in Weeks 3-4 are now out of the formula entirely. Their close wins in Weeks 5-6 are down to 20% strength. Everything else has a single-game DVOA of 30% or higher. The only weird thing about Baltimore's current ratings is that their weighted special teams rating is 22nd. That's because of one bad game, Week 15 against the Jets. Without that game, their special teams DVOA goes up to 11th in weighted ratings and fifth for the entire season.

The clear outlier among the remaining playoff teams is Houston, way down at 23rd in weighted DVOA. That would be 21st without counting Week 17, when they sat starters against Tennessee, but it's not a huge change. The Texans have three big losses on their resume, with a brutal -83% DVOA for their Week 11 loss to Baltimore. That's even with the opponent adjustments; that game would be -121% without opponent adjustments. They also had big losses to Denver and to Tennessee (with the starters on the bench). Five of their last six wins have been by less than a touchdown. Two of those wins have negative DVOA (Oakland Week 8, 27-24, and Indianapolis Week 12, 20-17) and two others (Weeks 15-16 over Tennessee and Tampa Bay) are very close to zero. The win over Kansas City is now a long time ago. It's still counting a little bit in the weighted DVOA formula, but actually Kansas City (25%) had a higher DVOA than Houston (7%) despite losing 31-24 when these first teams first met. It was one of those strange games where the winning team was less efficient but won by running may more plays than the other team. Houston had 83 plays to 47 for Kansas City, but the Chiefs had 6.6 yards per play compared to 5.7 for Houston. Over the long term, the superior yards per play is more of a positive indicator than running a lot of plays in the game, even if running a lot plays in that game leads to a win.

We'll get to another game like that when we review the single-game DVOA ratings for the wild-card round below.

* * * * *

To save people some time, we remind everyone to put their angry troll hatred into the official zlionsfan angry troll hatred Mad Libs form:

<team> is clearly ranked <too high/too low> because <reason unrelated to DVOA>. <subjective ranking system> is way better than this. <unrelated team-supporting or -denigrating comment, preferably with poor spelling and/or chat-acceptable spelling>

If you are new to our website, you can read the explanation of how DVOA is figured here. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

There are no adjustments here for sitting starters in Week 17, although we do adjust the ratings that we use in the playoff odds report.

You will find DVOA matchup pages for the four divisional games on the FO Premium page. Snap counts should be updated with information on the wild-card round by this evening.

Teams in yellow are still alive in the playoffs. Teams in gray lost this past weekend.

RK TEAM WEI.
DVOA
LAST
WEEK
W-L WEI OFF
DVOA
RK WEI DEF
DVOA
RK WEI S.T.
DVOA
RK TOTAL
DVOA
RK
1 BAL 55.7% 1 14-2 33.7% 1 -22.6% 2 -0.6% 22 41.5% 1
2 NO 41.0% 2 13-4 31.0% 2 -6.0% 11 4.0% 6 30.0% 3
3 KC 36.1% 3 12-4 20.4% 4 -7.4% 9 8.3% 1 30.2% 2
4 TEN 26.9% 5 10-7 28.7% 3 0.0% 17 -1.8% 25 10.8% 9
5 SF 24.1% 4 13-3 6.9% 8 -14.0% 4 3.2% 9 27.9% 4
6 NE 16.1% 6 12-5 -0.8% 15 -13.3% 5 3.7% 7 27.9% 5
7 MIN 14.4% 7 11-6 1.5% 12 -11.6% 6 1.3% 14 14.8% 7
8 DAL 12.5% 8 8-8 20.2% 5 3.0% 22 -4.7% 30 17.1% 6
9 SEA 10.8% 9 12-5 11.1% 6 1.4% 19 1.2% 15 13.4% 8
10 TB 7.9% 11 7-9 -7.9% 24 -19.5% 3 -3.7% 29 1.5% 14
11 BUF 7.0% 10 10-7 -5.7% 21 -11.2% 7 1.4% 13 1.9% 13
12 GB 5.3% 13 13-3 3.9% 9 -0.6% 16 0.7% 17 7.7% 10
13 LAR 5.2% 14 9-7 -0.3% 14 -7.8% 8 -2.4% 27 5.5% 11
14 ARI 3.8% 15 5-10-1 8.5% 7 2.9% 21 -1.8% 26 -5.8% 20
15 PHI 3.1% 12 9-8 -4.1% 19 -5.9% 12 1.2% 16 5.1% 12
16 ATL 1.9% 16 7-9 1.5% 13 -1.1% 15 -0.6% 23 -5.0% 17
17 CHI -4.4% 17 8-8 -9.9% 25 -5.6% 13 -0.1% 20 -2.1% 15
18 IND -5.4% 22 7-9 -6.9% 22 1.7% 20 3.3% 8 -5.0% 16
19 LAC -6.3% 18 5-11 2.9% 10 1.3% 18 -7.9% 32 -6.7% 21
20 NYJ -8.4% 19 7-9 -19.9% 30 -7.0% 10 4.4% 4 -15.8% 26
21 PIT -9.5% 20 8-8 -36.1% 32 -23.8% 1 2.8% 11 -5.5% 18
22 DEN -9.8% 21 7-9 -13.7% 26 -3.4% 14 0.5% 19 -9.0% 22
23 HOU -12.1% 23 11-6 -3.6% 18 13.6% 27 5.2% 3 -5.7% 19
24 CLE -15.0% 24 6-10 -1.8% 17 12.5% 26 -0.7% 24 -10.1% 23
25 OAK -16.3% 25 7-9 2.4% 11 16.2% 28 -2.6% 28 -11.0% 24
26 NYG -16.4% 26 4-12 -7.4% 23 11.1% 25 2.1% 12 -17.5% 27
27 CIN -18.7% 27 2-14 -15.5% 27 9.5% 24 6.2% 2 -25.2% 29
28 MIA -21.2% 29 5-11 -1.7% 16 19.1% 32 -0.4% 21 -36.7% 32
29 DET -22.2% 28 3-12-1 -5.5% 20 17.2% 29 0.5% 18 -12.1% 25
30 WAS -23.7% 30 3-13 -19.1% 29 8.9% 23 4.2% 5 -25.7% 30
31 JAX -30.3% 31 6-10 -16.3% 28 17.2% 30 3.2% 10 -18.0% 28
32 CAR -47.5% 32 5-11 -22.1% 31 17.6% 31 -7.9% 31 -26.9% 31

Here are the single-game DVOA ratings for the first round of the playoffs. The Tennessee Titans end up with a much higher DVOA than the New England Patriots despite a close game, but a lot of that is the effect of the final pick-six. (Throwing a short interception when backed up into your own end zone is very bad.) The surprise here is New Orleans with a much higher DVOA rating than Minnesota. This is another game where the winning team ran more plays but the losing team was more efficient. The Saints had 6.0 yards per play compared to 4.9 for Minnesota, although the Saints did lose the turnover battle. But Minnesota ran 74 plays to just 54 for New Orleans. (That advantage was 65 to 54 in regulation, and then the Saints never saw the ball in overtime.) Minnesota also won with much better third-down efficiency, which doesn't necessarily translate long term but certainly helps win a single game:

MIN-NO Offensive DVOA by Down, Wild Card
Down MIN NO
1 -22.0% 38.9%
2 -3.5% 54.1%
3 53.7% -42.0%
ALL 3.7% 24.4%

And of course, New Orleans screwed up their time management, taking 40 seconds off the clock, wasting the two-minute warning, and leaving themselves with a timeout at the end of regulation that they never used.

Here are all the single-game ratings.

DVOA (with opponent adjustments)
TEAM TOT OFF DEF ST
BUF -10% -15% -3% 3%
HOU -5% -4% 6% 6%
TEN 44% 21% -19% 5%
NE -5% -14% -3% 6%
MIN 6% 4% -9% -7%
NO 41% 24% -11% 6%
SEA 4% -3% -8% 0%
PHI -22% -32% -1% 9%
VOA (no opponent adjustments)
TEAM TOT OFF DEF ST
BUF -9% -9% 3% 3%
HOU -5% -10% 0% 6%
TEN 25% 2% -17% 5%
NE -10% -11% 4% 6%
MIN -26% -5% 14% -7%
NO 16% 8% -2% 6%
SEA 2% -4% -6% 0%
PHI -28% -25% 12% 9%

* * * * *

Once again this season, we have teamed up with EA Sports to bring Football Outsiders-branded player content to Madden 20 on a monthly basis. Today, we get to announce the Football Outsiders December players for Madden Ultimate Team on consoles, which will go live at 10:30am Eastern on Sunday. (We chose these players based on play in Weeks 14-17, leaving out December games that were in Week 13.)

  • HB Kenyan Drake, ARI: Led all running backs in rushing DYAR in December (69 carries, 400 yards, 7 TD).
  • WR Breshad Perriman, TB: Led all wide receivers in receiving DYAR in December (20-for-31, 419 yards, 5 TD).
  • TE Noah Fant, DEN: Third among tight ends in receiving DYAR in December (9-for-11, 183 yards, TD).
  • C Matt Paradis, CAR: Carolina third in adjusted line yards on runs up the middle, only two blown blocks in December.
  • RG Graham Glasgow, DET: Detroit tenth in adjusted line yards on runs up the middle, only one blown block in December.
  • ROLB Harold Landry, TEN: Tied for eighth with 12 hurries in December.
  • MLB James Burgess, NYJ: Led NFL with 13 defeats in December. Average run tackle for just 1.5 yards, lowest of any MLB.
  • CB J.C. Jackson, NE: 4.9 yards per pass in coverage in December according to Sports Info Solutions charting.
  • SS Julian Love, NYG: Second in NFL with 9 defeats in December. Average run tackle after just 2.1 yards and allowed just 3.6 yards per pass in coverge.
  • SS Eric Reid, CAR: Led all defensive backs with 24 run tackles in December.
  • K Randy Bullock, CIN: 10-for-10 in December including six field goals of 44 or more yards.
  • P Logan Cooke, JAX: Led NFL with 4.9 points of estimated field position on net punting in December; league-leading 50.1 gross yards per punt.

Comments

40 comments, Last at 08 Jan 2020, 8:51pm

1 Minn - NO

so take away OT and NO would have even more dominant DVOA. Sometimes DVOA doesn't fit with what I was watching. This game was one of those. 

30 Aaron - if you are watching,…

Aaron - if you are watching, I have a question about third downs.

I thought one of the FO premises was that third down performance reverts to the mean over the long run.  I found this article from 2010 that mentions it: https://www.footballoutsiders.com/extra-points/2010/stat-day-third-down-rebound-effect

 

The article references the "Football Outsiders Basics" page.  But I can't find that tenant on the page any more.  Has the third down rebound premise been debunked?

34 Over the last few years, we…

Over the last few years, we've found a lot less third-down regression on the offensive side of the ball than we used to see. So we removed it from the FO basics page for now.

One thing we keep meaning to do is an offseason project of testing all those FO basics things so we get updated numbers to quote and can determine if any of them need to be removed/edited.

3 I'd like to see a quarter…

I'd like to see a quarter and score analysis of Zimmer teams with the lead. He's a terrific coach, but he's always been too willing, for my comfort, to trade yards for clock in the 2nd half with a lead, especially by 8 or more. It almost always works out fine, and when Rhodes was hurt yesterday, his hand may have been forced to an even greater degree, to play soft, but the Vikings lose a lot of DVOA in the 4th quarter of wins, relative to other teams. At least that's my impression.

7 I felt the same thing but I…

I felt the same thing but I read somewhere Zimmer has never lost a game for Minnesota in which his team led by 10 pts. 

Of interest - in the 1st half the

1st down runs 9 for 58 yards.

1st down passes 6 for 4 net yards (includes -6 Diggs trick play ruled sack and def int)

2nd Half

1st down runs 11 for a net of 0 yards (includes 10 yard off holding)

1st down passes 4, completing 2 gaining 61 yards (includes 10 yard def pi penalty)

 

 

 

 

23 That pitch play got shut down in the second half

While I think it is true he chooses to run more in the 2nd half with a lead, it is also predictability. That pitch play was called maybe 4-5 times in the second half and it was no gain or a loss every time. They even called it at the goal line on the gaming winning drive. NO knew it was coming. Predictability for Zimmer in the second half I think is a greater problem than simply running more often.

8 Don't have time today to go…

Don't have time today to go back and check years past, but in 2019, Minnesota's offense rises from 4.6% to 10.3% in the second half with 8+ leads, and their defense stays more or less put, going from -9.9% to -9.5%.

It should be noted, of course, that games where you have leads in the second half are usually games you're playing well to begin with, so you'd expect DVOAs to be a little higher in general, at least offensively -- defenses around the league do get a little bit softer when defending leads late, as time becomes a bigger problem than points or yards.

 

 

24 Right, I mean beyond the…

Right, I mean beyond the obvious that they blew the 10 point lead, New Orleans had a pretty clear read on Minnesota's 1st and 2nd down RB run plays in the 2nd half. Cook took his first carry of the 3rd quarter for 11, and then between him and Mattison they went for 3, 1, 1, 1 (TD), -2, 3, -1, -7, -2 before overtime.

The Diggs jet sweeps broke things up nicely though. That's one thing I wish more teams would do in these types of situations. (Looking at you too, Tennessee - they were very, very lucky to hold on to that one point lead.) I'm all on the side of those who advocate for more passing and play action in general, but if you do want to run the ball and you do want to run out the clock at the end of games, you don't have to run the same plays and slam your RB into loaded boxes. Truly prolific rushing offenses at the lower levels understand this very well - they use all sorts of motion, misdirection, and the QB (when possible) to present lots of different looks and keep defenses on their heels.

28 Completely agree. This…

Completely agree. This always baffles me about running out the clock. I mean I see a lot of 3rd-and-2 fake-one-way, pitch-the-other misdirection plays, but they always seem to be used in the same situation. What if you did it on first down?

10 Belichick worst coached game of the year?

Was it the worst coached game of the year by any coach in any game? Belichick had 4 decisions that cost 32.1 GWC combined. How much GWC did NE have to begin with? I saw a press conference during the season where Belichick scoffed at analytics, I thought that he was just being Belichick and that in the playoffs he would change and just go for it, and that we would see a different NE team.

14 Having trouble scrolling down to read article

In reply to by Aaron Schatz

Is anyone else having this problem? 

I am not surprised that Anthony Lynn had the worst coached game of the year.  In the playoff game last year on 3rd and a mile very late in the 4th quarter and the Ravens having no timeouts, he threw a pass completed over the middle 10-15 yards down the field, well short of a first down, leading to a punt on the next play.   An incomplete pass would have given the Ravens extra time which they may have used to make the improbable comeback.   

15 don't take models too seriously

I didn't like the decision to punt late in the game.  But as for earlier kick/punt decisions: it's a bit nitpicky to isolate a decision that "cost" the Patriots 0.9% according to some model that may or may not be relevant to the game at hand.  It's important to understand that these models are just models, and don't represent actual probabilities.  They should be interesting when the probabilities cited are large (such as the 18.3% chance on the last punt) but not remotely interesting when small.  There's zero chance that this model is calibrated to take into account all the individual particulars of the game in question between these two teams on this particular day, at least not to the extent that we can take seriously any claim that kicking a field goal "cost" the Patriots 0.9% win percentage.

If you have a model that is systematically overestimating a team's odds of converting a first down, then it may well end up calling good decisions bad ones.  

Having watched the game and seen the Patriots get stuffed three times consecutively at the 1 yard line, I'd be very suspicious of any model that seems to think any kind of punt or kick is a bad idea.  Said model is probably not tuned to this particular game.

I didn't like the final NE punt because I thought too little time was left on the clock and because at that point field position mattered far less than possession.  By which I mean, if the Titans converted two first downs, they would win the game, no matter where on the field that happened.

I am fine with the other punts and kicks because, let's face it, the offense was dreadful.  

I have more issues with the offensive playcalling than these kick/punt decisions.  And I have more isues with the execution of plays and with the penalties than I do with the playcalling.  Shaq Mason bizarrely being downfield removed a long reception that would have put the Patriots in field goal range.  

17 Yep, there is no such thing…

Yep, there is no such thing as an "on average" football game. There are only particular games, with particular athletes. That isn't to say any coach should just ignore what happens on average, but it can't be automatically applied, either.

(edit) Yeah that Mason penalty was really weird, like he microdosed some acid, briefly hallucinated Lamar Jackson was his qb, thus went to throw a block downfield, then the acid wore off, and he remembered that his qb was born when Fran Tarkenton was still playing.

19 I was surprised that…

I was surprised that Belichick's repeated poor calls on 4th down was hardly commented on in Audibles at the Line. I think coaches like Jason Garrett or Andy Reid would've been buried alive for those calls. Belichick seems to have a halo effect which protects him from criticism even when it is deserved. I think the Patriots likely win if he was more aggressive.

20 The first call I didn't like…

The first call I didn't like was to kick the FG to go up 13-6 after being stuffed 3 times on the goal line on three consecutive runs.

I think you run your best 2-point conversion play there to try and score on 4th down.  If you fail, you still have TEN backed up inside their 5 and a decent chance to get the ball back before halftime on a short field that might still let you kick a (likely longer) FG.

The punts near midfield in the second half were bad, too.  NE's D was playing well enough and their O poorly enough the gamble to extend the drives seemed like the better bet.

Overall, it looked like a bad game for Belichick, other than the defensive adjustments at halftime. 

26 Yeah, it's not wrong to…

Yeah, it's not wrong to point out that the models have limitations in that they don't consider the teams/units on the field, but it's not hard to flip the usual arguments on their head like you do here. It seems like a lot of people are defending the midfield punts by saying that New England has a poor offense that is less likely than average to convert 4th down, and a defense that is likely to get a relatively quick stop. Why can't those same facts be used to justify being more aggressive on 4th down? If New England's offense is bad, can we be sure that they'll have many more opportunities with this good of field position in a close game? If their defense is good, doesn't that limit the downside potential of failing to convert?

35 I didn’t mind the second…

I didn’t mind the second half punts, because by that stage the NE defense was dominating, and (I’m aware this sounds incredibly anti-analytical) it felt as though a blunder from Tannehill was on its way. Keeping Tennessee pinned deep and awaiting a blunder didn’t feel like bad strategy in a knife-edge game with their own offence spluttering. But perhaps my instincts are also too conservative. 

(As it was, a blunder did indeed arrive when Tannehill fumbled the snap, but they got away with it. Then he converted a couple of huge third downs, and the punter did the rest.)

13 NE-Tenn DVOA

I am surprised to see that the interception at the end of the game counted so much toward DVOA. I figured that this would be calculated as an incomplete as if it were an intercepted hail mary pass on the final play of the game or half. In this situation a 98 yard play where time expired was also worthless, technically the interception returned for a TD was better, it gave New England the chance at music city miracle Part 2, plus a two point conversion for victory or a PAT for overtime. Any way of seeing DVOA discounting the pick six? Watching the game I felt that Belichick cost the Patriots the game (see my prior comment)

16 agree

In reply to by jheidelberg

DVOA might want to take game situation into account a bit more.  With that little time left on the clock, any kind of play failure is equally bad.  If Brady had found, say, Phillip Dorsett open in the middle of the field for a 55-yard gain, the Patriots would have also lost the game because they would have run out of time.  

 

18 DVOA is not win percentage,…

In reply to by RickD

DVOA is not win percentage, but also, down 1 in a game where DPI has an infinite yardage potential is still sub-optimal.

39 DVOA isn't trying to…

In reply to by RickD

DVOA isn't trying to evaluate games, it's evaluating *teams.* A team that's able to throw a 55-yard pass in the middle of the field against the Titans defense completed the kind of play that tends to lead to points. Points tend to lead to wins. That's the connection from DVOA to wins.

If you're trying to rate someone's performance by game winning potential, I don't see how you do that, because you can't isolate offense/defense anymore because you're coupled by the game itself.

It's just EPA vs WPA. The EPA for that play was like, -6.62, I think, whereas the WPA for that play has got to be something like -1% or so (unless you're PFR, apparently, who thinks that being down by 1 on your own 1 with 15 seconds left is a 10% chance of winning).

21 Oddly, I sort of agree with…

In reply to by jheidelberg

Oddly, I sort of agree with DVOA on this play, not because the pick 6 mattered, but because how the heck was that play call even a thing?

Ryan returning the ball for the TD was actually better for NE than if Sanu had caught it.

Heave it down field and hope a Titan DB has a brain glitch and runs over somebody seemed like pretty much the only change NE had at that point. 

22 MINN vs NO

I get the theory of DVOA and that plays are weighted based on value gained over the long term, but the Vikings were clearly the better team in this game. Take away two big plays from Hill and the NO offense didn't have anything but short plays that were bottled up. The clock management snafu I don't think actually changed the win percentage significantly. Brees just can't throw deep and so having an extra 40 seconds wouldn't have made the win any more likely.

DVOA is a calculation so its not biased but the narrative people want is the NO screwed it up. The reality is the Vikings played the better game and the NFC is stacked so its anyone's conference. If you think otherwise you were watching a different game.

Note: the Vikings made it a lot closer by continuously calling the pitch plays in the second half that got shut down. Stefanski needs to adjust and not be so predictable.

25 I am a bit surprised the…

In reply to by stevo

I am a bit surprised the Saints finished in the positive on offense and that the Vikings needed opponent adjustments to be in the black on defense. But based on Quick Reads I bet the Taysom Hill plays played a big role in that. Otherwise I definitely think they had the "normal" Saints offense pretty well under control.

31 Reality Vs. Narrative

In reply to by stevo

Reality is that Brees had 8 passes of at least 14 yds, so it wasn't all short passes being bottled up. Even without Hill's 100 yards on 1 pass and four runs, they had 224 yards on 49 plays, which is right around the same as the Vikings per play average, and the Hill plays count.

Reality is that the Vikings punted on 4 of 5 possessions in the second half and only picked up 2 first downs on those four punt possessions. They also went backwards on one of the possessions in which they picked up a first down, and ended up with -3 yards on six plays for the possession. 

The Vikings had an excellent plan. They were able to get sporadic pressure on the Saints, disrupting their methodical offense. They also benefited from a tremendously boneheaded play by Brees trying to throw a 50 yard air pass - something that he has not been able to do for a really long time, and got lucky when Ginn fell down and was too lazy to get up to tag Harris down. 

On the other hand, the Saints got lucky with the Thielen fumble and were able to benefit from the unexpected pass by Hill. 

The upshot of all of this was a 20-17 game with the Saints facing a first down at the 20. At that point, the Vikings got very lucky on the Brees fumble, then even luckier when the Cook fumble was called back, then lucky again when Kamara inexplicably lunged forward at the moment of the spike, which completely changed the end of the game, from the Saints having a chance to win in regulation to basically needing to play for the field goal, which they did poorly. 

Then the Vikings got lucky in winning the coin toss. Then they played well on offense for the first time since the third quarter, benefited from the Saints #1 corner getting rolled up on to hit a long pass play and won the game.

If you want to say that narrative of a game that was 20-20 in regulation is that one team dominated, you can, but you should at least have some numbers that back it up.

33 Agree, partially

I don't take much issue with the Vikings DVOA on offensive. They were sporadically effective and in the second half, beyond the touchdown drive, went conservative and couldn't nail the game down. And I don't think the Vikings dominated, but they did play the better game overall.

I do take issue with the Saints DVOA on offense and the Vikings DVOA on Defense. The Vikings got the better of the Saints offense overall. Maybe this is caused by the relative extremes of the Saints play. Meaning, when the Saints offense was bad, they had few plays per drive lowering their overall impact. And when they were good, they had long drives or big plays and scored without much in between. Although I have not looked at the numbers on that.

I also think the word luck needs to be used carefully. The assumption is "all fumbles are lucky" however many fumbles are 1) intentionally initiated by the defender, for example by a direct punch to the ball or 2) caused directly by pressure to the QB like what happend to Drew Brees. If that fumble is lucky, then the pressure on him also has to be considered lucky which is quite a rabbit hole to go down (If a statistical study is done, I would assume teams that cause more pressure create more QB fumbles). I know DVOA makes assumptions based on "on average, all fumbles are assigned this much luck and therefore have a reduced value" but that is a weakness of DVOA when applied to a single game which has its own very specific set of circumstances which do not necessarily represent the average value.

Ultimately, I like DVOA a lot, I just think in individual game, it has weaknesses but over a longer period of time, these weaknesses become less apparent. For example, I take no issue with the Vikings season long DVOA. I think it is pretty accurate.

37 It's not that DVOA thinks…

In reply to by stevo

It's not that DVOA thinks fumbles are lucky--in fact, quite the opposite.  There are DVOA big penalties for fumbling and big ones for causing fumbles.  What DVOA thinks are lucky are fumble recoveries.

32 How was the Brees fumble …

How was the Brees fumble "very lucky" for the Vikings? Hunter makes an athletically superior play to beat an excellent tackle, then makes an athletically superior play to whack Brees hard enough on the tricep to cause Brees' grip on the ball to fail, while tackling Brees, thus limiting his ability to recover his fumble? Similarly, the failure to get lined up right is not "luck", as I understand the term.I wouldn't call Thielen's fumble luck either, except in a somewhat minor sense, that receiver fumbles near the sideline often bounce out of bounds. The Cook nonfumble wasn't luck. He landed hard on his knee, then his elbow landed hard on something. There wasn't a fumble. That's not luck, as I understand the term. The winning td may be considered a minor instance of luck, in that some refs certainly would have have called offensive pi, which certainly would not have been overturned, either. That kind of either way call isn't really uncommon however. To me, the term luck is better employed in games like the last Saints/Vikings playoff game in New Orleans, when the game had about 9 total fumbles, and the Saints recovered about 6 or 7 of them, or when the refs failed to call multiple instances of Favre being roughed. The fact that the Vikings defense dropped multiple easy interceptions in that game was not an instance of the Saints being lucky, but rather just an instance of the Saints' opponent failing in basic execution, like Favre making a bad decision on his last pass. I thought this most recent game was closely matched, with Zimmer getting a jump on Payton in the 1st half, then both teams having some bad luck with injuries to important cornerbacks in the 2nd half that featured in the scoring, while Payton also got the jump on Zimmer at times throughout the game, with the Taysom Hill stuff that the Vikings weren't ready for. It seems obvious to me that the most critical unforseen event was Thielen and Cook returning to healthy athleticism for the first time in about 8 weeks.

40 AFC Super Bowl odds seem high

Adding up the Super Bowl odds for the 4 AFC teams is 68%. That feels really high and implies Baltimore is probably like 75%. I would take a NFC-wins-Super-Bowl at +200 in a heartbeat.